In the aftermath of yesterday's hung parliament Italian vote, we said that at least in retrospect, JPM had made it very clear that should the current outcome materialize then at least according to the house of Morgan, one should sell Italian bonds. Sure enough, the same analyst Gianluca Salford just said to go "risk off" on BTPs.
Update after Italian vote: switch to risk-off portfolio
- The Italian elections provided a massive surprise and resulted in hung Parliament
- The strong political message is a rejection of austerity measures
- We believe there will be an attempt to form a grand coalition (70%), but see a risk of centre-left/centre coalition being enlarged by senators defecting from the centre-right or Grillo’s party (15%), or another election (15%)
- We turn tactically risk-off and recommend outright longs in 10Y Bunds, 5s/10s German curve flatteners and shorts in 10Y Italy
- LTRO repayments from peripheral banks are likely to slow down: close EONIA steepeners, while keeping long Jun13 ECB OIS
- Initiate Bund swap spread wideners and stay long gamma: buy 3Mx10Y straddles
- FX: EUR/USD falls to the high 1.20s in an adverse scenario of sovereign spread widening and German 2Y rates returning to crisis levels of 0%. Undershoots have been common during the sovereign crisis, though this Italian episode is much less existential than last year’s Greek elections. Positions look neutral now by two measures (IMMs and currency manager betas).
In line with 2006 and 2008 elections, polls, exit polls and instant polls all failed to give a correct indication of voting patterns, overestimating support for the Bersani’s centre-left and underestimating support for Berlusconi’s centre-right, but especially Grillo’s 5 Stars movement (Exhibit 1). In terms of seat allocation (Exhibit 2), in the Chamber (lower house) the centre-left managed to beat the centre-right by less than 1% and got the majority of seats. However it managed to win only eight of the seventeen regions due to the plurality vote system in the Senate (upper house). Getting a majority in the Senate is therefore very complicated. We see two negatives stemming from this vote:
1) The political message coming from the popular vote is a rejection of austerity measures, with the parties more in tune with the austerity programme suffering the most. Monti’s centre components underperformed expectations by the most in relative terms and barely exceeded the 10% threshold in the Chamber. The performance of Grillo’s party was impressive, gaining almost 10% compared to the pre-blackout polls, and also Berlusconi’s centre-right comeback from the December lows is worth highlighting. As discussed in previous emails (What happens if Berlusconi wins?, Alex White, 4 February and A Crash Course in Grillonomics, Alex White, 21 February), the centre-right coalition and the 5 Stars movement share a deep scepticism about the Euro area reform agenda.
Exhibit 1: The Italian elections provided a massive surprise…
Vote % in lower and upper house and versus latest polls;
Exhibit 2: … and resulted in a hung Parliament
Seat allocation in lower and upper house; seats and %
2) There is high political uncertainty as the agreement on how to get a workable majority in both houses is not straightforward and another round of elections in the near term, although unlikely, cannot be excluded. In Italy, both houses have essentially the same role and the Prime Minister needs a majority from both. We see three scenarios:
i) In our view the most likely one will be an attempt to form a grand coalition government which includes centre-left, centre and centre-right (70% probability). We do not expect the three coalitions to agree on much more than a new electoral law, and a modicum of fiscal tightening to keep a benign fiscal outlook (Italy’s primary surplus is already satisfactory). However, we would expect little progress on much needed structural reforms. The most important force holding these coalitions together is likely to be the fear of a Grillo landslide if new elections are held. At the moment it is unclear who might be the PM, given Monti’s electoral defeat.
ii) An alternative to an explicit grand coalition is for a centre-left plus centre government to reach a majority in the senate thanks to defections from the centre-right or Grillo’s party. Given that the number of senators needed is high, around 15-20, we do not attach more than 15% probability to such an outcome, most likely after the failure to reach an agreement on a grand coalition.
iii) Finally, early elections are a possibility (15% probability). However, the time required to hold a second vote is long (around two months), and we expect market and political pressure to avoid the uncertainty; as discussed above a new election in the short term is likely to see an increase in support for the centre-right but especially for Grillo’s 5 Stars movement. It is interesting to highlight that in last year’s Greek second election, support converged to the most anti-austerity party among the mainstream ones (New Democracy) and to the most anti-austerity party among radical ones (Syriza) (Exhibit 3).
Exhibit 3: We expect fear of an early election to force a grand coalition between centre-left, centre and centre-right; last year in Greece, anti-austerity vote increased in the second election
% of total votes won by the major political parties in the May 2012 and June 2012 Greek elections;
The market implications are not positive in our view: we see risks of no agreement or slow progress on a grand coalition over the next few days. Even if an agreement is reached we see a very weak political mandate for further austerity measures and any type of structural reforms. This coupled with recent weakness in some macro releases, is likely to halt the progress on the virtuous circle of improving financial conditions, lower volatility and increasing investor appetite for riskier assets such as peripheral bonds. Although we believe tail risk is greatly reduced relative to last summer, we recommend investors to open risk-off trades. Technicals are supportive, with our client survey showing that benchmarked investors entered the Italian elections long peripherals vs. core countries (Exhibit 4).
Exhibit 4: Benchmarked investors entered the Italian elections long peripherals
Net exposure of Euro area investors to peripheral vs. core countries; %
We recommend longs in 10Y Bunds (with a 1.35% target) and find 5s/10s flatteners an attractive bullish proxy. We unwind trades with a bearish duration bias such as 3s/7s steepeners and 10s/30s flatteners in Germany. In terms of core spreads, we close 5Y overweights in Belgium and turn neutral. In peripherals, we open shorts in 10Y Italy as we believe that 10Y BTP yield could exceed 5.00% in coming days.
The increased political uncertainty in Italy is also likely to slow down the LTRO repayments from peripheral banks, with Italian banks expected to be particularly slow in paying back. We revise upward our estimate of excess liquidity by year end to €300bn (from €250bn) and we tactically close front/greens and front/golds Eonia steepeners, while keeping longs in Jun13 ECB OIS.
Our bullish outlook on 10Y German yields combined with higher excess liquidity, higher swaption volatility and likely light swapped issuance activity makes us tactically bullish on swap spreads. We open Jun13 Bund swap spread wideners with a target of 38bp by the end of March.
Political uncertainty is expected to increase delivered volatility, while keeping implieds well bid. We keep a bullish outlook on volatility and in addition to long Jun green midcurve straddle we recommend buying gamma in 3Mx10Y. We take profit in long 3Mx5Y gamma vs. 3Mx20Y.
The EUR/USD forecast of 1.32 in Q1 and Q2 has always incorporated two risks: that the Euro area’s exit from recession could prove protracted and uneven, and that political risk would rise in late February/early March due to Italian elections and the unclear reform momentum of even a centre-left or grand coalition. As a reminder of how much the return of sovereign stress could weaken EUR/USD, every 50bp of spread widening in the periphery (based on 5Y rates) tends to motivate a 2-cent decline, as does every 10bp fall in ECB rate expectations versus the Fed. Thus in an adverse scenario in which sovereign spreads widen a total of 100bp and 2Y German yields return to mini-crisis levels near 0%, EUR/USD would fall about 8 cents, so to the high 1.20s from EUR/USD’s February peak. Note that some of this re-pricing has already occurred in the run-up to and immediate aftermath of elections, since the German 2Y has fallen from around 0.25% to 0.09%, average 5Y spreads have widened by about 35bp and the currency has fallen almost 6 cents this month.
The risk is for a euro undershoot, as has occurred during every previous period of sovereign stress, such as Greek elections in summer 2012 when the currency declined about 5 cents more than justified by moves in sovereign spreads . Greece merited a decent risk premium last year because the country's EMU membership was in doubt unless it elected a centrist government. By contrast in Italy the risks are far less existential – they center on the possibility of fiscal slippage and a lack of structural reform momentum under a grand but fragile coalition. Still, the broader point is that as long as the composition and policies of the next Italian government persist, the euro should underperform versus most currencies except where some local factor like the prospect of further debt monetization remains an almost equally powerful issue, as in the UK (GBP) and Japan (JPY). Note that there is no short base to contain the euro’s decline, since most position measures like the IMMs and currency manager betas are close to neutral rather than underweight.
What’s next? We expect politicians to start discussing the option of a grand coalition as soon as possible. After this morning’s T-bill auction, Italy will sell €4.75-6.50bn of bonds including a new 10Y tomorrow. The election of the presidents of the lower house and upper house should be held in the second half of March and this might be the first test of a new grand coalition.